ABSTRACT
The present chapter is concerned with the philosophical tradition of understanding causes as raising the probability of their effects. It discusses the probability theories of causality of Suppes and Granger and introduces Zellner’s idea of using causal laws to determine the relevance of the variables and lags to be included in a model representing relations of Granger causality. It also discusses causal Bayes nets theory and emphasizes that knowledge of causes that raise the probability of their effects can be employed for purposes of prediction, but less so for purposes of policy analysis. It finally mentions a number of problems that are potentially inherent to attempts to infer causality from probabilities when causality is understood in accordance with another philosophical tradition (that of understanding causes as causally dependent on instrumental variables).
